Sample Size Re-estimation (SSR)

Resources

Statistical software for design and monitoring

 Sample size estimation and SSR methods based on interim estimates of nuisance parameters can be easily implemented using routine statistical software although there are also a few more specialised resources available.

Open-access software


spass (R package)
Allows sample size estimation and blinded SSR in full population or subpopulation.

blindConfidence (R code on github)
Simulation code for exploring the impact of blinded SSR on estimation. See underlying methods for details.   

 BayesianSSRwithPDCCPP (R code on github)
 Presents code for illustrating Bayesian SSR using power priors. See underlying methods for for details.

gsDesign (R package)
The varBinomial function computes blinded estimates of the variance of the estimate of: 1) event rate differences, 2) logarithm of the risk ratio, or 3) logarithm of the odds ratio. This is intended for blinded SSR for comparative trials with a binary outcome. The ssrCP function offers comparative SSR based on conditional power which has not been covered in this section. 

Close source software


tba

Commercial software

nQuery
 Performs blinded and unblinded SSR for binary and continuous outcomes with restricted and unrestricted decision rules.

SSR considerations and guidance for practice


  • Pritchett et al. Sample size re-estimation designs in confirmatory clinical trials—current state, statistical considerations, and practical guidance. Stat Biopharm Res. 2015;7(4):309–21.